undefined | Better HN

0 pointsrayiner1d ago0 comments

It depends what you’re using the data for. If you’re comparing across countries, or looking at a developing country over time, it’s a relatively small factor. The ratio between the average and the median isn’t that big even in the U.S. (about 1.3). Meanwhile, Poland’s GDP per capita has tripled since 2005: https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?location....

0 comments

mejutoco1d ago

> The ratio between the average and the median isn’t that big even in the U.S. (about 1.3)

Being off by 30% might not matter for some usages, but it is not a small amount. It seems the median is more accurate to report and we agree.

dhosek1d ago

I don’t know that there is any way to calculate a median value of GDP/capita. You can look at income distributions and find a median income and compare that to a mean income, which could allow an estimation, but beyond that, GDP is an intrinsically composite number which cannot be easily (at all?) broken down to individual contributions. I assume income is what the parent commenter is basing the median-mean comparison on, but it’s kind of out of nowhere with no explanation.

rayinerOP1d ago

For purposes of comparing countries to each other and the same country over time, it’s not 30% off. The average skews higher than the median everywhere, so it’ll be 1.3/1.x.

If you have reported median incomes that are calculated the same way across countries and over time, that would be better. But many countries don’t reliably track that data, and the ones that do calculate it in completely different ways.

mejutoco23h ago

That all sounds reasonable. My concern was with your quote

> GDP/capita aligns very closely with material standard of living for the median person

GDP per capita is an average. This means it does not align with the _median_ person, but with the average. I believe this is factual and undeniable. No doubt it is interesting too to try to find other metrics for different usages as well.

> For purposes of comparing countries to each other and the same country over time, it’s not 30% off

You said the ratio for the same country between gdp mediand and average is 1.30. That means it is 30% off. Again, we can keep moving the goalposts and I could agree, but for the quoted statements i believe the above is true.

1 more reply

j / k navigate · click thread line to collapse